GESMES XML
GESMES XML
The aim of our work is to implement a generic software package which will be able to automatically:
i) Extract statistical information from a database / datawarehouse (or any other source) using metadata and definitions;
ii) Format this information into GESMES/XML messages
iii) Deliver this information over the Internet in encrypted and compressed form.
iv) Receive and parse GESMES/XML messages;
v) Allow viewing and validation of the data before its storage; and
vi) Store the data in the target datawarehouse.
The status of the development can be summarised as follows:
- Design of an XML specification for GESMES, mostly based on (and compliant to) the existing GESMES/CB XML proposal. Note that while GESMES/CB is mostly confined to time-series data, our GESMES/XML proposal is suited for generalised multidimensional statistical messages. Similarly, development of a GESMES query message (i.e. a message requesting GESMES data).
- Development of machine-readable XML schemas for the two messages (similar to XML DTDs), which can be used to validate and interpret relevant GESMES / XML messages.
- Development of a GESMES/XML parsing library in both C++ and Java, which can be used in any software package requiring parsing and interpreting of GESMES/XML messages.
- Design and prototype implementation of a generic metadata repository able to hold GESMES structural definitions for specific cube structures as well as logical links to the physical tables of a database with the actual data. This generic metadata repository built closely following a UML model of GESMES, can be used to provide definitions to an application aiming to automatically extract data from a statistical database and generate a suitable GESMES/XML message.
- Development of a web-accessible prototype application which is able to receive XML query mesaages, parse and translate them into SQL statements, retrieve the data, format it into GESMES response messages and deliver it.
- Development of a prototype Java GESMES/XML viewer which allows the user to slice and view the multidimensional cube in a message.
We are planning to develop an EDI-to-XML translator for GESMES in order to ensure interoperability with older applications. Furthermore, the design of the software has the flexibility to accomodate any XML specification, so that there will be no problem when a definitive XML standard for GESMES become official.
Geographical/ Statistical Database Design [up]
At the core of the STATLAS system will be its database where all geographical, cartographic and statistical information will be stored. The organisation of this database is tailored to the overall design of the system. In particular the statistical part of the database will contain the pre-processed entities in the appropriate format. Data will be collected, evaluated, organised into thematic categories, pre-processed and stored in the database. The database will be open in the sense that it will be horizontally and vertically expandable and updateable. This is a critical characteristic of the database, in view of the fact that the system will address the data "on the fly" in order to execute the statistical processing and the subsequent portrayal. This will allow for future addition of other entities, updating of the database with recent data and flexibility in the utilisation of its content. This design that incorporates the multidimensional structure of Statistical data with the object relational geographical into a flexible and versatile database is expected to be usable in data situations other than STATLAS as well.
The geodatabase is the repository of spatial data inside a DBMS. It contains vector data, raster data, tables, and other GIS objects. A significant task in designing a geodatabase involves defining the entities in the geodatabase and how they will be represented as features, rasters, attributes, and so on. In many, if not most, GIS implementations, it is important to have the ability to define spatial data as a system of objects and relationships. Indeed, part of the power of GIS is the ability to model and represent the relationships among these objects.
STATLAS database has been implemented in the Oracle environment. Oracle Spatial, often referred to as Spatial, provides a SQL schema and functions that facilitate the storage, retrieval, update, and query of collections of spatial features in an Oracle8i database.The Geographic data stored in the STATLAS geo- database has been derived from GISCO, that is the GIS of the European Commission, created and maintained by Eurostat. Information covers the administrative regions, altimetry, hydrography, infrastructure, land and natural resources in Europe, and the world database.
